Landscape ecology deals with the effects of the spatial configuration
of mosaics on a wide variety of ecological phenomena. Because problems
in many areas of conservation biology and resource management are rel
ated to landscape use, development of a rigorous theoretical and empir
ical foundation for landscape ecology is essential. We present an appr
oach to research that focuses on how individual-level mechanisms opera
ting in a heterogeneous mosaic produce ecological patterns that are sp
atially dependent. The theoretical framework that we develop considers
the density and distribution of a population among patches as a funct
ion of (a) within-patch movement patterns of individuals; (b) emigrati
on from patches as a function of population density, patch configurati
on, patch context, and within-patch movement; and (c) loss of individu
als as they disperse through landscape elements. This theoretical fram
ework is coupled with an empirical approach that emphasizes the use of
experimental model systems (EMS), small-scale systems occupying ''mic
rolandscapes.'' Such systems are amenable to experimental manipulation
with adequate replication and control and may serve as analogs of sys
tems occurring at broader landscape scales. where rigorous empirical w
ork is more difficult. We illustrate the use of EMS with examples from
our studies of beetles and ants in North America and of voles in Norw
ay. Finally, we consider some factors that may constrain extrapolation
s from EMS to other systems, scales, or levels. We believe that implem
entation of a mechanistic approach to landscape ecology is essential t
o deriving generalizations about how spatial heterogeneity influences
ecological systems.